Pléiades Neo Super Resolution

This algorithm improves the spatial resolution of Pléiades Neo image from 0.3 m to 0.1 m (by factor of 3).


Description

A processing block that improves the spatial resolution of a Pléiades Neo image from 0.3 m to 0.1 m by using a trained convolutional neural network. To use this block, request access.

See this block on the marketplace.

How it works

This processing block only expresses features that are shown in the original image.

It takes around 16 minutes to enhance a 3.92 km2 image. To avoid long processing time, you can choose to enhance only one band combination, RGB or NED.

ParameterOverview
RGBboolean / required if
Required if NED=false.
An RGB band (red, green, and blue) will be enhanced.
NEDboolean / required if
Required if RGB=false.
A NED band (near-infrared, red edge, and deep-blue) will be enhanced.

Compatibility

Catalog datasets

Use in combination with Processing from Storage.

Example

A workflow with the following blocks:

{
  "up42-processing-from-storage:1": {
    "asset_ids": ["c569b4f2-08c3-5cj6-b6ob-fee52dbfec12"]
  },
  "data-conversion-dimap:1": {
    "ms": true,
    "pan": false,
    "bbox": null,
    "contains": null,
    "intersects": null,
    "clip_to_aoi": false
  },
  "pleiadesneo-superresolution:1": {
    "NED": false,
    "RGB": true
  }
}

Capabilities

Input

raster
up42_standard
bands
[
  "red",
  "green",
  "blue",
  "nir",
  "rededge",
  "deepblue"
]
dtype
{
  "or": [
    "uint16",
    "uint8"
  ]
}
formatGTiff
sensor
{
  "or": [
    "Pleiades",
    "PNEO"
  ]
}

Output

raster
up42_standard
bands
{
  "or": [
    [
      "red",
      "green",
      "blue"
    ],
    [
      "nir",
      "rededge",
      "deepblue"
    ]
  ]
}
dtypeuint8
format> (propagated)
sensor> (propagated)
To know more please check the block capabilities specifications.